ML Anthology

Estimating Relatedness via Data Compression

Brendan Juba
ICML 2006 pp. 441-448
doi:10.1145/1143844.1143900 /icml/2006/juba2006icml-estimating/

Abstract

We show that it is possible to use data compression on independently obtained hypotheses from various tasks to algorithmically provide guarantees that the tasks are sufficiently related to benefit from multitask learning. We give uniform bounds in terms of the empirical average error for the true average error of the n hypotheses provided by deterministic learning algorithms drawing independent samples from a set of n unknown computable task distributions over finite sets.

PDF ICML Semantic Scholar

Cite

Text

Juba. "Estimating Relatedness via Data Compression." International Conference on Machine Learning, 2006. doi:10.1145/1143844.1143900

Markdown

[Juba. "Estimating Relatedness via Data Compression." International Conference on Machine Learning, 2006.](https://mlanthology.org/icml/2006/juba2006icml-estimating/) doi:10.1145/1143844.1143900

BibTeX

@inproceedings{juba2006icml-estimating,
  title     = {{Estimating Relatedness via Data Compression}},
  author    = {Juba, Brendan},
  booktitle = {International Conference on Machine Learning},
  year      = {2006},
  pages     = {441-448},
  doi       = {10.1145/1143844.1143900},
  url       = {https://mlanthology.org/icml/2006/juba2006icml-estimating/}
}